基于高斯混合模型的火焰检测算法  

Flame detection algorithm based on Gaussian mixture model

作　　者：张怡[1] ZHANG Yi(The Engineering&Technical College of Chengdu University of Technology,Leshan 614000,Sichuan Province,China)

机构地区：[1]成都理工大学工程技术学院,四川乐山614000

出　　处：《信息技术》2021年第1期74-79,共6页Information Technology

基　　金：四川省教育厅科技项目(17ZB0062);乐山市科技局项目(19SZD106)。

摘　　要：火灾作为对社会和环境危害最大的灾难,一直是人们重点防范的对象。但目前现有的火灾预警系统都存在误报率过高的问题。因此,文中提出了一种基于火焰闪烁动力学的火焰检测框架。在该框架中,火焰颜色分布模型采用高斯混合模型。此外,采用概率显著性分析方法和一维小波变换提取运动显著性和滤波后的时间序列作为特征,描述火焰的动态特性和闪烁特性。通过实验证明了提出的方法对比现有方法具有较好的精确度,能够获得95%以上的精度,并且具有较低的误报率,满足实际需求。As the most harmful disaster to society and the environment,fire has always been the focus of attention.However,the existing fire early warning systems all have the problem of high false alarm rate.Therefore,a flame detection framework based on flame flicker dynamics is proposed.In this framework,the flame color distribution model adopts the Gaussian mixture model.In addition,the probabilistic saliency analysis method and one-dimensional wavelet transform are used to extract motion saliency and filtered time series as features to describe the dynamic characteristics and flicker characteristics of the flame.Experiments have proved that the proposed method has better accuracy compared with the existing methods,can obtain an accuracy of more than 95%,and has a lower false alarm rate,which meets actual needs.

关 键 词：机器视觉 火焰检测 高斯混合模型 Dirichlet过程 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]